Mesogeos: A multi-purpose dataset for data-driven wildfire modeling in the Mediterranean

We introduce Mesogeos, a large-scale multi-purpose dataset for wildfire modeling in the Mediterranean. Mesogeos integrates variables representing wildfire drivers (meteorology, vegetation, human activity) and historical records of wildfire ignitions and burned areas for 17 years (2006-2022). It is designed as a cloud-friendly spatio-temporal dataset, namely a datacube, harmonizing all variables in a grid of 1km x 1km x 1-day resolution. The datacube structure offers opportunities to assess machine learning (ML) usage in various wildfire modeling tasks. We extract two ML-ready datasets that establish distinct tracks to demonstrate this potential: (1) short-term wildfire danger forecasting and (2) final burned area estimation given the point of ignition. We define appropriate metrics and baselines to evaluate the performance of models in each track. By publishing the datacube, along with the code to create the ML datasets and models, we encourage the community to foster the implementation of additional tracks for mitigating the increasing threat of wildfires in the Mediterranean.

[1]  Ari S. Morcos,et al.  A Cookbook of Self-Supervised Learning , 2023, ArXiv.

[2]  I. Papoutsis,et al.  Deep Learning for Global Wildfire Forecasting , 2022, ArXiv.

[3]  I. Papoutsis,et al.  Wildfire Danger Prediction and Understanding With Deep Learning , 2022, Geophysical Research Letters.

[4]  V. Kotroni,et al.  Meteorological Analysis of the 2021 Extreme Wildfires in Greece: Lessons Learned and Implications for Early Warning of the Potential for Pyroconvection , 2022, Atmosphere.

[5]  Yi-Fan Chen,et al.  Next Day Wildfire Spread: A Machine Learning Dataset to Predict Wildfire Spreading From Remote-Sensing Data , 2021, IEEE Transactions on Geoscience and Remote Sensing.

[6]  Nuno Carvalhais,et al.  Deep Learning Methods for Daily Wildfire Danger Forecasting , 2021, ArXiv.

[7]  J. Thepaut,et al.  Supplementary material to "ERA5-Land: A state-of-the-art global reanalysis dataset for land applications" , 2021, Earth System Science Data.

[8]  Deep Learning for the Earth Sciences , 2021 .

[9]  J. Keeley,et al.  Wildfires and global change , 2021 .

[10]  Siyuan Ma,et al.  Gated Transformer Networks for Multivariate Time Series Classification , 2021, ArXiv.

[11]  R. Lily Hu,et al.  Convolutional LSTM Neural Networks for Modeling Wildland Fire Dynamics , 2020, ArXiv.

[12]  Ross D. Shachter,et al.  Uncertainty Aware Wildfire Management , 2020, ArXiv.

[13]  Yi-Fan Chen,et al.  Deep Learning Models for Predicting Wildfires from Historical Remote-Sensing Data , 2020, ArXiv.

[14]  E. Chuvieco,et al.  Fire Danger Observed from Space , 2020, Surveys in Geophysics.

[15]  Mark Crowley,et al.  A review of machine learning applications in wildfire science and management , 2020, Environmental Reviews.

[16]  J. Pereira,et al.  Wildfire management in Mediterranean-type regions: paradigm change needed , 2020, Environmental Research Letters.

[17]  J. San-Miguel-Ayanz,et al.  A global wildfire dataset for the analysis of fire regimes and fire behaviour , 2019, Scientific Data.

[18]  Brian Y. Lattimer,et al.  Wildland Fire Spread Modeling Using Convolutional Neural Networks , 2019, Fire Technology.

[19]  Markus Reichstein,et al.  Earth system data cubes unravel global multivariate dynamics , 2019, Earth System Dynamics.

[20]  Ming Wang,et al.  Forest Fire Susceptibility Modeling Using a Convolutional Neural Network for Yunnan Province of China , 2019, International Journal of Disaster Risk Science.

[21]  P.J. Denning,et al.  On learning how to predict , 1980, Proceedings of the IEEE.

[22]  David Radke,et al.  FireCast: Leveraging Deep Learning to Predict Wildfire Spread , 2019, IJCAI.

[23]  Chen Chen,et al.  SmokeNet: Satellite Smoke Scene Detection Using Convolutional Neural Network with Spatial and Channel-Wise Attention , 2019, Remote. Sens..

[24]  Quoc V. Le,et al.  EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks , 2019, ICML.

[25]  J. Randerson,et al.  The Global Fire Atlas of individual fire size, duration, speed and direction , 2018, Earth System Science Data.

[26]  Younes Oulad Sayad,et al.  Predictive modeling of wildfires: A new dataset and machine learning approach , 2019, Fire Safety Journal.

[27]  Joachim Denzler,et al.  Deep learning and process understanding for data-driven Earth system science , 2019, Nature.

[28]  J. Randerson,et al.  The Global Fire Atlas of individual fire size, duration, speed, and direction , 2018 .

[29]  T. Dube,et al.  Accuracy assessment of MODIS active fire products in southern African savannah woodlands , 2018 .

[30]  P. Ciais,et al.  FRY, a global database of fire patch functional traits derived from space-borne burned area products , 2018, Scientific Data.

[31]  Moulay A. Akhloufi,et al.  Computer vision for wildfire research: An evolving image dataset for processing and analysis , 2017 .

[32]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[33]  C. Cammalleri,et al.  Comparing soil moisture anomalies from multiple independent sources over different regions across the globe , 2017 .

[34]  Stephan Hoyer,et al.  xarray: N-D labeled arrays and datasets in Python , 2017 .

[35]  J. Keeley,et al.  Flammability as an ecological and evolutionary driver , 2017 .

[36]  Andrew J. Tatem,et al.  WorldPop, open data for spatial demography , 2017, Scientific Data.

[37]  Philippe Ciais,et al.  The status and challenge of global fire modelling , 2016 .

[38]  C. Justice,et al.  The collection 6 MODIS active fire detection algorithm and fire products , 2016, Remote sensing of environment.

[39]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[40]  M. Moritz,et al.  Climate change‐induced shifts in fire for Mediterranean ecosystems , 2013 .

[41]  F. Ballester,et al.  Health effects of the 2012 Valencia (Spain) wildfires on children in a cohort study , 2013, Environmental Geochemistry and Health.

[42]  J. San-Miguel-Ayanz,et al.  Analysis of large fires in European Mediterranean landscapes: Lessons learned and perspectives , 2013 .

[43]  Kirk R. Klausmeyer,et al.  Climate Change, Habitat Loss, Protected Areas and the Climate Adaptation Potential of Species in Mediterranean Ecosystems Worldwide , 2009, PloS one.

[44]  Christopher I. Roos,et al.  Fire in the Earth System , 2009, Science.

[45]  D. Roy,et al.  What limits fire? An examination of drivers of burnt area in Southern Africa , 2009 .

[46]  Peter Vogt,et al.  European Forest Fire Information System (EFFIS)—rapid damage assessment: Appraisal of burnt area maps in southern Europe using MODIS data (2003–2005) , 2006 .

[47]  A. Scott,et al.  The diversification of Paleozoic fire systems and fluctuations in atmospheric oxygen concentration. , 2006, Proceedings of the National Academy of Sciences of the United States of America.

[48]  W. Bond,et al.  Fire as a global 'herbivore': the ecology and evolution of flammable ecosystems. , 2005, Trends in ecology & evolution.

[49]  S. Hochreiter,et al.  Long Short-Term Memory , 1997, Neural Computation.

[50]  P. Fiorucci,et al.  Wildfire hazard mapping in the eastern Mediterranean landscape , 2023, International Journal of Wildland Fire.

[51]  Ross D. Shachter,et al.  WildfireDB: An Open-Source Dataset Connecting Wildfire Spread with Relevant Determinants , 2021 .

[52]  Jesús San-Miguel-Ayanz,et al.  Towards a coherent forest fire information system in Europe: the European Forest Fire Information System (EFFIS). , 2002 .

[53]  M. Finney.  FARSITE : Fire Area Simulator : model development and evaluation , 1998 .

[54]  P. Andrews.  BEHAVE : Fire Behavior Prediction and Fuel Modeling System - BURN Subsystem, Part 1 , 1986 .

[55]  R. Burgan,et al.  BEHAVE : Fire Behavior Prediction and Fuel Modeling System -- FUEL Subsystem , 1984 .